Invasive Procedures for Cancer Pain
Between 70% and 90% of all cancer pain can be controlled with oral medication1, but for those patients with unrelieved pain invasive procedures have an important role. Recent surveys show that (1) cancer pain is both underdiagnosed and underrated and (2) physicians unfamiliar with current pain treatment modalities are more likely to support assisted suicide for their patients2. Appropriate use of invasive measures in the 10ñ30% of patientsómost often those with advanced diseaseówho fail oral therapy can relieve nearly all cancer pain.
It is important to emphasize that regional techniques such as nerve blocks for cancer pain management are intended to be analgesic "adjuvants" and not definitive treatment. These procedures should allow patients to lower drug dosages and thereby reduce side effects, or to experience better pain relief from current dosages in order to improve their quality of life. Neither primary physician nor pain specialist should promise permanent relief, since the patientís disease may progress and spread. Twycross reported that most patients referred for cancer-related symptom management have at least two anatomically distinct pain sites, and more than 40% have four or more sites3. It is unreasonable to expect one regional block to eliminate pain from its multiple sites. On the contrary, medical care of the suffering pain patient requires a multimodal, multispecialty approach combining psychotherapy, social support, and pain management to provide the best possible quality of life or quality of dying4.
Application of invasive measures to the 10-30% of patients who fail oral therapy can relieve nearly all cancer pain.
Invasive techniques for managing cancer pain often employ neurolytic substances like ethanol or phenol. A thorough knowledge of relevant anatomy and the mechanism of action by which the agent destroys neural tissue is essential to minimize irreversible complications. Several points must be addressed before one proceeds with a neurolytic block. Only physicians with extensive experience and skill should perform these blocks. As stated above, nerve blocks should be regarded as part of a multimodal approach to pain and not as a stand-alone pain "cure." Patients should be thoroughly informed about likely sensory deficits and possible complications. In most cases, neurolytic blocks should first be simulated with local anesthetic to allow the patient to experience the sensory changes that may occur5. The patient should be followed for several days after the diagnostic block. Finally, close monitoring and planned opioid reduction should follow a successful neurolytic block to prepare for somnolence and respiratory depression when the respiratory stimulation of pain is removed.
Neurolytic Agents
Ethanol (alcohol). Ethanol has been used extensively for neurolytic procedures in concentrations from 3% to 100%. It acts by destroying nerves and producing Wallerian degeneration without disruption of the Schwann cell sheath. Thus axonal regeneration can and will occur, sometimes resulting in neuroma formation. However, if cell body destruction occurs, regeneration is not possible. Recent studies have shown that ethanol destroys nervous tissue by extraction of cholesterol and other lipids and by protein precipitation. Topical application of alcohol to exposed nerves results in both axon and Schwann cell destruction6.
Studies in the early part of this century demonstrated that approximately a 50% alcohol solution is required for analgesia when performing major nerve blocks. Labat showed that 33% alcohol can produce analgesia without motor paralysis when applied to peripheral nerves. When alcohol is applied to autonomic nerves and ganglia, all effector organ input is blocked. Permanent blocks occur when postsynaptic nerves in the ganglia are affected. Blockade limited to the rami or preganglionic fibers produces only temporary analgesia followed by return of sensation in three to six months. Subarachnoid injection of 100% alcohol produces damage to the dorsal roots, Lissauerís tract, and posterior columns. Wallerian degeneration follows nerve destruction. Almost 90% of injected alcohol is removed by the cerebrospinal fluid (CSF) within 10 minutes after subarachnoid injection.
Clinically, alcohol in a concentration of 33ñ100% is hypobaric (specific gravity = 0.8) relative to CSF (s.g. = 1.1), so target sites in the spinal cord should be above the site of injection to allow the alcohol to float upward. The effective duration of intrathecal alcohol analgesia is about six months. Complete, prolonged analgesia can be achieved if cell bodies are destroyed along with axons in peripheral and autonomic nervous tissue. High concentrations of alcohol (90ñ100%) sometimes produce a chemical neuritis under clinical conditions, probably owing to residual partial neural destruction of some tissues. Injection of alcohol into peripheral nerves or ganglia not surrounded by fluid requires no special positioning.
Phenol. Studies by Mandl in 19507 reported that 6% phenol applied to cervical ganglia in animals produced local necrosis in 24 hours, complete degeneration by 45 days, and regeneration in 75 days. Thus, sensory recovery after phenol is faster than after alcohol. Phenol, like alcohol, has been administered for subarachnoid, peripheral nerve, and ganglion neurolysis.
In the subarachnoid space, phenol is hyperbaric with respect to CSF. The target site should be below the site of injection to allow the phenol to migrate downward. Concentrations of 5ñ10% produce nonspecific neural destruction similar to alcohol. Some studies suggest that larger diameter fibers are damaged to a greater extent than smaller diameter fibers. Extensive fibrosis and thickening of the arachnoid can also occur. Phenol destroys axons along the dorsal roots and posterior columns without disrupting cell bodies.
When 3ñ6% phenol is applied to peripheral nerves, both acute and chronic damage to axons and myelin occur. In addition, protein coagulation and necrosis ultimately result in axonal and Wallerian degeneration. Phenol is injected in a range of concentrations from 3% to 15%. Suggestions that phenol may be more destructive to vascular than neural tissue have not been borne out clinically. Lema et al. reported that autopsy samples of patients who had received 3.4 grams of phenol interpleurally showed little evidence of vascular, organ, or even neural histological destruction8.
Invasive Techniques
To be considered invasive, a procedure must violate the integument of the body. Thus, all needle and scalpel procedures are invasive. Moreover, patient-controlled analgesia utilizing intravenous, subcutaneous, or epidural sites may be considered invasive in this sense. Invasive procedures can be categorized according to three general targets: autonomic nervous system, peripheral nerve, and neuraxis.
Autonomic Nervous System Blocks
The autonomic nervous system is largely responsible for visceral nociception. A diagnostic local anesthetic block of a sympathetic nerve or plexus establishes the relative contribution of autonomic and visceral pain and can simulate the effect of neurolytic nerve block (Table 1).
|
Table 1. AUTONOMIC NERVE BLOCKS |
|
|
Neurolytic Block |
Site/Condition Treated |
|
Stellate ganglion |
Head or arm pain |
|
Gasserian ganglion |
Trigeminal neuralgia and facial pain |
|
Interpleural (thoracic sympathetic chain) |
Upperóhead, arms Middleóthorax, heart, lung Loweróbladder, abdominal organs, uterus |
|
Celiac plexus (splanchnic nerves) |
Pancreatitis, abdominal pain, visceral cancer pain |
|
Lumbar sympathetic |
Lower limb pain |
|
Hypogastric plexus |
Perineal, pelvic, and lower limb pain |
|
Sacrococcygeal ganglion (impar, Walther) |
Rectal pain |
Stellate ganglion block. The stellate ganglion lies anterior to the lateral process of the C7 vertebra. Autonomic pathways to the ipsilateral head and upper extremity are interrupted by a block of this ganglion. Because of the proximity of other vital structures, many clinicians are reluctant to perform neurolytic blocks in this area. However, serial blocks with neurolytic agents in dilute concentrations (3ñ6% phenol) after encouraging diagnostic local anesthetic block have been recommended. Potential complications include injection into the vertebral artery, phrenic and superior laryngeal nerve block, and rarely, intrathecal injection.
Celiac plexus block. The celiac plexus lies on the anterolateral surface of the aorta at the T12 to L2 vertebral level. Blockade of this plexus reduces visceral pain from abdominal organs and has gained widespread acceptance for the treatment of pancreatic cancer pain9. The incidence of pain relief has been reported to be more than 84%, although occasionally repeat blocks are required10. Possible complications include transient hypotension; intrathecal, epidural, or intrapsoas injection; injection into the aorta or vena cava, puncture of the kidney, intestine, or lung; anterior spinal artery syndrome; paraplegia; and death.
Hypogastric plexus block. The hypogastric plexus lies anterior to the L5 to S1 vertebrae and controls autonomic activity to the pelvis and lower limbs. A block of these nerves has been described to reduce pain associated with pelvic malignancy11. Injury to sacral nerves, bladder or bowel perforation, intravascular injection, and urinary or fecal incontinence are potential complications.
Ganglion impar. Intractable perineal pain presents problems to the pain practitioner because somatic, visceral, and autonomic nerves controlling excretory and sexual functions converge in the pelvis. Blockade of the ganglion impar (also called the ganglion of Walther or sacrococcygeal ganglion) provides pain relief without significant somatovisceral dysfunction for many patients with advanced cancer. The ganglion impar, the only unpaired autonomic ganglion, lies anterior to the sacrococcygeal junction and can be blocked with 5ñ10 cc of solution administered through a needle penetrating the anococcygeal ligament or directly passing through the sacrococcygeal (calcified) ligament12. Patients experiencing significant rectal discomfort or pain after abdominoperineal resection of rectal cancer often benefit from this technique.
Interpleural analgesia. Interpleural analgesia has been successfully used for minor surgical anesthesia and in the management of certain chronic pain states such as pancreatic pain and post-thoracotomy pain syndrome. It has also been used to alleviate acute exacerbations of pain in advanced stages of various cancers.
The mechanism of action of interpleural analgesia appears to be somatic nerve block, but an autonomic block has also been suggested. Diffusion of local anesthetic into the brachial plexus and cervical sympathetic ganglion can relieve upper extremity pain while also producing a Hornerís syndrome. The success of this block depends on the correct positioning of the patient such that the solution gravitates to the appropriate paravertebral area. Possible complications include tension pneumothorax, pleural infection, and local anesthetic systemic toxicity from rapid tissue absorption. Interpleural administration of phenol has also been successful in the long-term relief of cancer pain8.
Peripheral Nerve Blocks
Neurolytic blockade of peripheral nerves can be effective as an analgesic adjuvant to oral therapy, but the long-term benefits of this approach are controversial. One concern is the transient nature of the neurolysis, and possible neuritis or deafferentation pain experienced as the block "wears off." This problem is generally avoided by selecting patients likely to succumb before development of the neuropathic pain (i.e., within four to six months). Common peripheral blocks and the conditions treated are listed in Table 2.
|
Table 2. PERIPHERAL NERVE BLOCKS |
|
|
Neurolytic Block |
Site/Condition Treated |
|
Ophthalmic |
Eye pain (glaucoma, uveitis) |
|
Maxillary, mandibular |
Tic douloureux or cancer pain |
|
Glossopharyngeal |
Tic-like jaw pain |
|
Phrenic nerve |
Hiccoughs, diaphragmatic pain |
|
Vagus, tracheo-bronchial |
Cancer of trachea |
|
Intercostal |
Thoracotomy scar pain, rib metastases |
|
Ilioinguinal/iliohypogastric |
Groin pain |
|
Sacral nerves |
Pelvic, rectal pain (alternative to spinal or caudal epidural blocks) |
Intraspinal Therapies
Intraspinal administration of opioids is frequently used to treat pain not controlled with oral medications5. Opioids can be delivered by the spinal or epidural routes to provide analgesia, generally at a lower dose than for systemic administration and without the motor, sensory, or sympathetic block associated with intraspinal local anesthetic administration13 (Table 3).
|
Table 3. NEURAXIAL BLOCKS |
|
|
Neurolytic Block |
Condition Treated |
|
Subarachnoid |
Unilateral segmental terminal cancer pain |
|
Epidural Neurolytic |
Same except used in non-terminal patient
|
Combinations of opioids plus local anesthetics given spinally produce effective analgesia with relatively few side effects. Spinal opioids can be delivered by intermittent bolus injection or by continuous infusion. Morphine is most commonly given, although hydromorphone, fentanyl, sufentanil, and oxymorphone have also been successfully used. Opioids in combination with clonidine are often effective in patients with advanced disease. However, sedation and hypotension associated with clonidine can limit titration14.
Three systems used for chronic intraspinal opioid administration include percutaneous tunneled epidural or spinal catheters, tunneled catheters connected to subcutaneously implanted injection ports, and implanted infusion pump systems15. Implantable pumps are more convenient to manage and less likely to become infected, and become cost-effective when life expectancy exceeds six months. Tolerance to chronic intraspinal administration of opioids is managed by increasing doses, changing to another opioid, or substituting local anesthetic for a short period. Other opioid side effects include pruritus, urinary retention, somnolence, myoclonus, catheter infection, and rarely, respiratory depression.
Neurosurgical Procedures
With the development of the multidisciplinary approach to pain management and an ever-growing range of available pharmacologic agents, few patients require surgical intervention to interrupt central or peripheral nociceptive pathways.
The most commonly performed surgical procedure for cancer pain relief is anterolateral cordotomy, which ablates the spinothalamic tract. This procedure can be done by an open technique, which has significant morbidity; potential complications include hemiparesis, urinary retention, and sexual impotence. Percutaneous cordotomy has largely replaced the open method and is usually performed under local anesthesia by advancing a thermal coagulation probe with fluoroscopic guidance. It is usually ineffective in neuropathic pain because it does not reverse central sensitization, and has only limited use in visceral pain. Immediate pain relief is achieved in the majority of patients, but pain recurs in roughly half of these by six to twelve months. Many patients in whom pain recurs also develop paresthesias or dysesthesias.
Placement of an Omaya reservoir under the scalp, connected to a catheter whose tip lies within the lateral cerebral ventricle, may provide satisfactory analgesia with relatively few side effects15,16. Other less often used procedures include rhizotomy, dorsal root entry zone (DREZ) lesioning, commissurotomy, and dorsal root ganglionectomy (Figure 1).
|
|
Figure 1. Cross-section of the spinal column showing various types of neurosurgical procedures (from Saberski and Ligham17).
Conclusion
For a minority of patients with pain due to cancer, oral drug therapy titrated according to the World Health Organization analgesic ladder fails to control pain adequately. In these patients, who often have an advanced stage of disease, more aggressive intervention is required. In a field replete with clinical anecdotes and uncontrolled case series, there is an urgent need for clinical trials and outcomes assessment to generate firm evidence on the risks, benefits, costs, and failures of such intervention. However, few would question that aggressive intervention is often appropriate. Invasive pain therapies must be provided with skill and caution, and are best offered within a multidisciplinary framework of care.
References
1. Zech DFJ, et al. Pain 1995; 63:65ñ76.
2. Cleeland CR, et al. New Engl J Med 1994; 330:552ñ596.
3. Twycross RG,
4. Alvarez Echeverri T, Velasquez Acosta O. Pain: Clinical Updates 1996; 4(2).
5. Cousins MJ, Bridenbaugh
ed.
6. Rumbsy MG, Finean JB. J Neurochem 1966; 13:1509ñ1513.
7. Mandl F. J Int Coll Surg 1950; 13:566ñ572.
8. Lema MJ, et al. Reg Anesth 1992; 17:166ñ170.
9. Eisenberg E, et al. Anesth Analg 1995; 80:290.
10. Brown DL, et al. Anesth Analg 1987; 66:869ñ873.
11. Plancarte R, et al. Reg Anesth 1997; 22:562ñ568.
12. deLeon-Casasola, OA. Tech Reg Anesth Pain Manage 1997; 1:27ñ31.
13. Plummer JL, et al. Pain 1991; 44:215ñ220.
14.
15. Ballantyne JC, et al. Reg Anesth 1996; 21:542.
16. Mummenthaler M, et al. (Eds). Treatment of Chronic Pain: Possibilities, Limitations, and
Long-term
Follow-Up.
17. Saberski L, Ligham D. Neuroablative techniques for cancer pain management. Techniques in
Regional Anesthesia and Pain Management 1997; 1(1):53ñ58.